The number of servers in a data center is usually enormous, from thousands, tens of thousands to hundreds of thousands or even more, and it is a problem that needs to be solved for each data center as to how to manage the servers in the data center, more particularly, how to determine the position of each server in the data center, and to know in time position changes of each server.
To solve this difficult problem, there are generally two solutions: one is to manually record the position of each server and changes of the position, and the other is to position each server by the radio frequency identification (RFID) technology. For the first solution, manual entry and registration is time consuming and has a high error rate, and it is not practical for the management of a large data center. The second solution may include two sub-solutions. According to the first sub-solution, an active tag may be installed on each server to actively send signals of a certain frequency, and an RFID reader may be installed on the rack of the data center to read and decode the information and then to send it to a central information system to perform related data processing. In the second sub-solution, an RFID tag is installed on the rack, the tag having stored therein the information of a corresponding physical position, and an RFID reader is installed on each server. After the server powers on, the reader will automatically scan the RFID tag to determine the position of the current server and store the result.
However, no matter what kind of the above RFID solutions is used, using the RFID technology to position servers can not realize accurate positioning, that is, since the RFID reader has a relatively remote reading distance, it can hardly know accurately the position of each server, rather, it can only know roughly the region where a server is. In addition, the implementation cost of the RFID technology is high, and it is very difficult to be applied in an actual application.